Cat Wu, who led Claude Code product, leaves Anthropic: "I never manually wrote a line of code" building an RL environment

Cat Wu, who led product for Claude Code, has left Anthropic. She wrote on X on Nov. 3, "Friday was my last day at Anthropic." The post passed 2,500 likes and 120,000 views.

Two years ago, when Claude Code was a small internal tool, she asked it to build a reinforcement-learning environment. She steered it for tens of hours without manually writing a line of code, it worked, and she joined the team. At launch most people approved every permission prompt by hand and mostly asked questions and wrote tests; today trusting Claude's judgment to auto-approve is the default, and Claude handles large migrations and new products, she wrote. She thanked the Claude Code, Cowork and Tag teams but named neither a next job nor a successor.

Colleagues including @trq212 posted farewells, and developer Addy Osmani wrote that she will be missed. The same day ClaudeDevs posted a recap, "October was a big month for Claude Code," covering mods, a better desktop app for running many sessions side by side, and more insight into usage (6,100 likes).

The official changelog shows /usage began showing large prompt-cache misses in version 2.1.311 on Oct. 26. Changelog versions run from 2.1.302 on Oct. 16 to 2.1.317 on Nov. 3.